Democratizing Machine Learning

The Data Exchange Podcast: Ameet Talwalkar on the Determined Training Platform, Neural Architecture Search, Federated Learning and more.


SubscribeiTunesAndroidSpotifyStitcherGoogle, and RSS.

In this episode of the Data Exchange I speak with Ameet Talwalkar, co-founder and Chief Scientist at Determined AI1, and an Assistant Professor in the Machine Learning Department at Carnegie Mellon University.  A few months ago, I spoke with one of Ameet’s co-founders (Evan Sparks), around the time they announced that they were open sourcing the Determined Training Platform (DTP).  Ameet and I started off by discussing the first few months of DTP as an open source project, specifically initial feedback from users, applications and use cases that they are seeing, and much more.

Scalable machine learning, scalable Python, for everyone: Join Michael Jordan, Manuela Veloso, Azalia Mirhoseini, Zoubin Ghahramani, Wes McKinney, Ion Stoica, Gaël Varoquaux, Raluca Popa, and many other speakers at the first Ray Summit, a FREE virtual conference which takes place Sep 30th and Oct 1st.

We then spoke at length about Ameet’s research interests and projects at CMU.  This includes:

  • Hyperparameter tuning and neural architecture search
  • Privacy-preserving tools, including federated learning
  • Ethics and fairness in machine learning
  • Democratizing machine learning
  • A video version of this conversation is available on our YouTube channel.

 
Subscribe to our Newsletter:
We also publish a popular newsletter where we share highlights from recent episodes, trends in AI / machine learning / data, and a collection of recommendations.

Related content:


[1] I am an advisor to Determined AI.

[Image: Abu Dhabi, Hyatt Capital Gate by Ben Lorica.]